Spine-Mets-CT-SEG | Spine metastatic bone cancer: pre and post radiotherapy CT

DOI: 10.7937/kh36-ds04 | Data Citation Required | Image Collection

Summary

Illustrative example dataset of a patient with vertebral lesions with the raw CT, annotated segmentation and identification of vertebral levels, and lesion classification (Lytic, Blastic and Mixed).

We provide an annotated imaging dataset of cancerous CT spines to help develop artificial intelligence frameworks for automatic vertebrae segmentation and classification. This collection contains a dataset of 55 CT scans collected on patients with a large range of primary cancers and corresponding bone metastatic lesions obtained for patients with metastatic spine disease. The subjects of the study planned for radiotherapy were simulated at the Radiation Oncology Department, Brigham and Women's Hospital, Boston, MA, using 1) Siemens SOMATOM Confidence (Siemens Healthcare GmbH, Erlangen, Germany) and 2) GE Lightspeed (General Electric Medical System, Waukesha, WI) CT scanner. Simulation scan parameters are detailed in Table 1.

 Table 1. CT image acquisition parameters.

*A: 16cm field of view, **B: Skin-to-Skin field of view 

The dataset includes: 

• Vertebral metastatic bone lesion classifications performed by RNA (20 years experience in biomechanics and image analysis of pathologic spines) and DBH (49 years experience in evaluating clinical imaging of pathologic spines) using standard radiological criteria as defined by the SINS protocol (J Clin Oncol, Aug 1;29(22):3072-7,.2011. doi:10.1200/ JCO. 2010 .34.3897). Specifically, vertebrae were classified as either Normal, Osteolytic, Osteosclerotic or having mixed lesions. 
• Manual segmentation labels for the vertebral levels was performed and verified by RNA in 3Dslicer (https://www.slicer.org/) 
• Deep-learning spinal segmentation model: Our automated segmentation model uses nnU-Net, a freely available open-source framework for deep learning in healthcare imaging, and is made publicly available. 
• Accompanying clinical data for each patient in the dataset includes demographics, primary cancer type lesion classification and vertebral involvement.

Data Access

Version 1: Updated 2024/09/25

Citations & Data Usage Policy

Data Citation Required: Users must abide by the TCIA Data Usage Policy and Restrictions. Attribution must include the following citation, including the Digital Object Identifier: